Progressive parallel interference canceller and method and receiver thereof

Abstract

A progressive parallel interference canceller (PPIC) and a method and a receiver thereof are illustrated. The PPIC reconstructs each subchannel interference reconstruction signal through several iterations and subtracts the corresponding subchannel interference reconstruction signal from each subchannel frequency-domain reception signal to obtain a subchannel frequency-domain signal. Thereby, according to the present disclosure, inter-channel interference can be cancelled without re-performing channel coding or estimating the signal to noise ratio (SNR) or frequency offset.

Description

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the priority benefit of Taiwan application serial no. 98124142, filed on Jul. 16, 2009. The entirety of the above-mentioned patent application is hereby incorporated by reference herein and made a part of this specification.

TECHNICAL FIELD

The present disclosure generally relates to a multiple input multiple output (MIMO) communication system, and more particularly, to a progressive parallel interference canceller (PPIC) of a MIMO communication system and a method and a receiver thereof.

BACKGROUND

Broadband transmission with high spectrum efficiency and high mobiling velocity (for example, high driving speed) is the essential requirement to future mobile communication systems. The multiple input multiple output (MIMO) technique along with the orthogonal frequency-division multiplexing (OFDM) modulation technique may be the most potential means for achieving foregoing requirement. However, the effect of inter-channel interference (ICI) will become very serious along with the increases of mobiling velocity, carrier frequency, and OFDM symbol time.

When the mobiling velocity exceeds 300 km/h, the signal interference noise ratio (SINR) of an entire communication system is limited by the ICI at 20 dB. As a result, the application of high-order modulation techniques (for example, 16-quardre amplitude modulation (16-QAM) or 64-quardre amplitude modulation (64-QAM)) is restricted. Regarding the future communication systems with high data rate and high mobiling velocity (for example, the Worldwide Interoperability for Microwave Access (WiMAX) system), the mobiling velocity may be up to 350 km/h or even 500 km/h. Accordingly, a MIMO detector and a subchannel interference canceller must be provided, such that a communication system can meet aforementioned requirement.

Presently, many communication equipment manufacturers and research organizations are dedicated to providing a subchannel interference canceller with high performance, high spectrum efficiency, and low complexity, so as to compensate the subchannel interference effect in the future communication systems (for example, the WiMAX system) with high data rate and high mobiling velocity. In addition, with such a subchannel interference canceller with high performance, high spectrum efficiency, and low complexity, a communication system may even adopt a high-order modulation technique, such as aforementioned 16-QAM and 64-QAM. Thereby, a subchannel interference canceller can be applied to any mobile communication system (for example, the WiMAX system and the Digital Video Broadcast-Terrestrial (DVB-T) system) based on the OFDM modulation technique to improve the market competitiveness of a communication electronic product and create a high industrial value.

SUMMARY

An exemplary embodiment of the present disclosure provides a progressive parallel interference canceller (PPIC) adaptable to a receiver in a multiple input multiple output (MIMO) communication system based on the orthogonal frequency-division multiplexing (OFDM) modulation technique, wherein the receiver performs a plurality of iterations. The PPIC includes a plurality of subchannel interference reconstruction units and a plurality of subchannel interference cancellation units. The k^th subchannel interference cancellation unit corresponding to an r^th receiving antenna is coupled to the k^th subchannel interference reconstruction unit corresponding to the r^th receiving antenna. The k^th subchannel interference reconstruction unit corresponding to the r^th receiving antenna generates a k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k+i−1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ during the t^th iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal Ŝ_t,q^(x) represents the q^th subchannel frequency-domain estimation signal corresponding to a t^th transmitting antenna generated during the x^th iteration, ((•))_N represents a calculation of dividing by N to obtain the remainder, N is the number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), and q=k−i+x or q=k+i−x. The k^th subchannel interference cancellation unit obtains a k^th subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna according to a k^th subchannel frequency-domain reception signal Y_r,k and the k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna during the i^th iteration.

An exemplary embodiment of the present disclosure provides a receiver adaptable to a MIMO communication system based on the OFDM modulation technique, wherein the receiver performs a plurality of iterations and includes a plurality of receiving antennas, a plurality of OFDM demodulators, a PPIC, and a MIMO detector. The PPIC includes a plurality of subchannel interference reconstruction units and a plurality of subchannel interference cancellation units. Each of the OFDM demodulators is coupled to one of the receiving antennas. The PPIC is coupled to the OFDM demodulators. The k^th subchannel interference cancellation unit corresponding to the r^th receiving antenna is coupled to the k^th subchannel interference reconstruction unit corresponding to the r^th receiving antenna. The MIMO detector is coupled to the PPIC. The k^th subchannel interference reconstruction unit corresponding to the r^th receiving antenna generates a k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k+i−1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ during the i^th iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal S_t,q^(x)represents the q^th subchannel frequency-domain estimation signal corresponding to a t^th transmitting antenna generated during the x^th iteration, ((•))_N represents a calculation of dividing by N to obtain the remainder, N is the number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), and q=k−i+x or q=k+i−x. The k^th subchannel interference cancellation unit corresponding to the r^th receiving antenna obtains a k^th subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna according to a k^th subchannel frequency-domain reception signal Y_r,k and the k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna during the i^th iteration. When the PPIC performs the i^th iteration, the MIMO detector determines a plurality of bits b₀, b₁, . . . , and b_MT−1 corresponding to the k^th subchannels of a plurality of transmitting antennas according to a plurality of k^th subchannel frequency-domain signals Y′_1,k⁽ⁱ⁾, Y′_2,k⁽ⁱ⁾, . . . , and Y′_R,k⁽ⁱ⁾ of the receiving antennas, wherein M is a bit modulation order, R is the number of the receiving antennas, T is the number of the transmitting antennas, and a plurality of bits b_M(t−1), b_M(t−1)+1, . . . , and b_Mt−1 corresponding to the k^th subchannel of the t^th transmitting antenna are used for determining the subchannel frequency-domain estimation signal Ŝ_t,k⁽ⁱ⁾.

An exemplary embodiment of the present disclosure provides a progressive parallel interference cancellation method adaptable to a receiver in a MIMO communication system based on the OFDM modulation technique, wherein the receiver performs a plurality of iterations. A k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to a r^th receiving antenna is generated according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k+i−1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ during the i^th iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal Ŝ_t,q^(x) represents the q^th subchannel frequency-domain estimation signal corresponding to a t^th transmitting antenna generated during the x^th iteration, ((•))_N represents a calculation of dividing by N to obtain the remainder, N is the number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), and q=k−i+x or q=k+i−x. Next, a k^th subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna is obtained according to a k^th subchannel frequency-domain reception signal Y_r,k and the k^th subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ corresponding to the r^th receiving antenna during the i^th iteration.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings are included to provide a further understanding of the present disclosure, and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

FIG. 1A is a graph illustrating the subchannel signal power distribution in a Worldwide Interoperability for Microwave Access (WiMAX) system.

FIG. 1B is a graph illustrating the subchannel interference power in a WiMax system.

FIG. 2 is a system block diagram of a receiver 300 in a multiple input multiple output (MIMO) communication system based on the orthogonal frequency-division multiplexing (OFDM) modulation technique according to an exemplary embodiment of the present disclosure.

FIG. 3 is a system block diagram illustrating part of the circuit of a progressive parallel interference canceller (PPIC) 340 according to an exemplary embodiment of the present disclosure.

FIGS. 4A˜4C are diagrams respectively illustrating how subchannel interference cancellation units 410˜412 and subchannel interference reconstruction units 420˜422 perform a 1^st to a 3^rd iterations.

FIG. 5 is a diagram of a message passing algorithm (MPA) according to an exemplary embodiment of the present disclosure.

FIG. 6 is a flowchart of a progressive parallel interference cancellation method according to an exemplary embodiment of the present disclosure.

DESCRIPTION OF THE EMBODIMENTS

Reference will now be made in detail to the present preferred embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers are used in the drawings and the description to refer to the same or like parts.

The present disclosure provides a progressive parallel interference canceller (PPIC) and a method and a receiver thereof, wherein each subchannel interference reconstruction signal is reconstructed through multiple iterations, and a corresponding subchannel interference reconstruction signal is subtracted from each subchannel frequency-domain reception signal to obtain a subchannel frequency-domain signal.

FIG. 1A is a graph illustrating the subchannel signal power distribution in a Worldwide Interoperability for Microwave Access (WiMAX) system, wherein the abscissa represents the subchannel index, and the ordinate represents the subchannel signal power (dB). Referring to FIG. 1A, the curve 102 is obtained at a mobiling velocity of 350 km/h (the product of the Doppler shift f_d and the orthogonal frequency-division multiplexing (OFDM) modulation symbol period T_s is 0.07), and the curve 101 is obtained at a mobiling velocity of 500 km/h (where f_d×T_s=0.1). As shown in FIG. 1A, the maximum value of the subchannel signal power falls around where the subchannel index is 0. Besides, as shown in FIG. 1A, the greater the mobiling velocity (i.e., f_d×T_s) is, the greater the subchannel signal power at where the subchannel indexes is not 0 is. Accordingly, the greater the mobiling velocity is, the more serious the subchannel interference is.

To be specific, when the mobiling velocity is 0, the signals only fall at where the subchannel index is 0. When the mobiling velocity increases, the signals are distributed to and become interference at where the subchannel index is not 0, and the interference increases along with the mobiling velocity. Serious subchannel interference with higher mobiling velocity may even cause a receiver to misunderstand the wireless signals transmitted by a transmitter and accordingly greatly reduce the system performance. Thus, the receiver has to estimate the magnitude of a subchannel interference and cancels the subchannel interference to reduce the probability of misunderstanding wireless signals transmitted by the transmitter.

FIG. 1B is a graph illustrating the subchannel interference power in a WiMax system, wherein the abscissa represents the value of f_d×T_s, and the ordinate represents the subchannel interference power (dB). Referring to FIG. 1B, the curve 201 represents an upper bound curve obtained through theoretical calculation, and the curve 202 represents a lower bound curve obtained through theoretical calculation. As shown in FIG. 1B, the greater the mobiling velocity (i.e., f_d×T_s) is, the more serious the subchannel interference is. Generally speaking, when the mobiling velocity exceeds 300 km/h, the signal interference noise ratio (SINR) is restricted at about 20 dB. Thus, the bit error rate (BER) of the entire communication system cannot be reduced no matter how great the signal to noise ratio (SNR) is.

The signal on a subchannel is mainly interfered by several adjacent subchannels. Accordingly, the present disclosure provides a PPIC. The PPIC cancels interference signals on each subchannel through iterations, and during an i^th iteration, the PPIC cancels the subchannel interference produced by signals on (2×i−2) adjacent subchannels. Below, the PPIC and a method and a receiver thereof provided by the present disclosure will be described in detail.

FIG. 2 is a system block diagram of a receiver 300 in a multiple input multiple output (MIMO) communication system based on the OFDM modulation technique according to an exemplary embodiment of the present disclosure. In the present exemplary embodiment, a transmitter in a MIMO communication system based on the OFDM modulation technique has T transmitting antennas, and the modulation method is to modulate M bits into a subchannel frequency-domain transmission signal S_t,k (i.e., the bit modulation order is M), wherein the subchannel frequency-domain transmission signal S_t,k represents a k^th subchannel frequency-domain transmission signal to be transmitted by the t^th transmitting antenna, t is an integer between 1 and T, k is an integer between 0 and (N−1), and N is the number of subchannels in the MIMO communication system.

Referring to FIG. 2, the receiver 300 includes a plurality of receiving antennas 391˜39R, a plurality of OFDM demodulators 311˜31R, a PPIC 340, a channel estimator 330, a multiplexer 320, a MIMO detector 350, and a channel decoder 360. The OFDM demodulator 31r includes a cyclic prefix (CP) remover 37r and a fast Fourier transformer (FFT) 38r, wherein r is an integer between 1 and R. The receiving antenna 39r is coupled to the CP remover 37r, and the CP remover 37r is coupled to the FFT 38r. The PPIC 340 is coupled to the OFDM demodulators 311˜31R, and the channel estimator 330 is coupled to the PPIC 340 and the OFDM demodulators 311˜31R. The multiplexer 320 is coupled to the MIMO detector 350, the channel decoder 360, and the PPIC 340, the MIMO detector 350 is coupled to the PPIC 340, and the channel decoder 360 is coupled to the MIMO detector 350.

A time-domain sampling signal S_t,u to be transmitted by a transmitter through the t^th transmitting antenna can be expressed as:

$S_{t,u}=\frac{1}{N}\sum _{k=0}^{N-1}{S}_{t,k}{e}^{j\frac{2\pi \mathrm{ku}}{N}}$ $u=0~N-1,t=1~T$
Next, the transmitter adds a CP to N time-domain sampling signals to be transmitted through the same transmitting antenna, wherein the u^th sample {tilde over (S)}_t,u of the CP in the OFDM symbol to be transmitted by the t^th transmitting antenna is expressed as:
{tilde over (S)}_t,u=S_t,N−NG+u 0≦u≦N_G−1,
wherein N_G represents the length of a guard interval. Finally, the t^th transmitting antenna transmits the OFDM symbol with the CP to the wireless channel.

The receiving antenna 39r receives the OFDM symbol transmitted by the transmitter from the wireless channel. The time-domain sampling signal in the OFDM symbol received by the receiving antenna 39r is a summation signal composed of the time-domain sampling signals transmitted by the transmitting antennas through the wireless channel. The CP remover 37r removes the CP in the OFDM symbol received by the receiving antenna 39r. The u^th time-domain sampling signal y_r,u corresponding to the r^th receiving antenna in the OFDM symbol after the CP is removed can be expressed as:

$y_{r,u}=\sum _{t=1}^T\sum _{l=0}^{N_G}{h}_{r,t,l}^{\left(u\right)}{s}_{t,{\left(\left(u-l\right)\right)}_N}+z_{r,u}0\le u\le N-1,1\le r\le R$
wherein h_r,t,l^(u) represents a l^th channel tap of a channel impulse response from the t^th transmitting antenna to the r^th receiving antenna at a u^th sampling time, ((•))_N represents a calculation of dividing by N to obtain the remainder, and z_r,u represents a u^th sample value corresponding to the additive white Gaussian noise (AWGN) of the r^th receiving antenna, wherein the average value and the variance of the AWGN are respectively 0 and σ_Z².

Next, the FFT 38r receives a plurality of time-domain sampling signals corresponding to the r^th receiving antenna and performs a fast Fourier transform to the time-domain sampling signals to generate a k^th subchannel frequency-domain reception signal Y_r,k corresponding to the r^th receiving antenna, wherein the subchannel frequency-domain reception signal Y_r,k can be expressed as:

$\begin{array}{c}{Y}_{r,k}=\sum _{u=0}^{N-1}{y}_{r,u}{e}^{-j\frac{2\pi \mathrm{ku}}{N_c}}0\le k\le N-1\\ =\sum _{t=1}^T{H}_{r,t,k}^0{S}_{t,k}+\underset{\underset{\mathrm{ICI}\mathrm{term}}{︸}}{\sum _{t=1}^T\sum _{d=1}^{N-1}{H}_{r,t,k}^d{S}_{t,{\left(\left(k-d\right)\right)}_N}}+Z_{r,k}\end{array}$
wherein Z_r,k represents the AWGN sample value of a k^th subchannel corresponding to the r^th receiving antenna. There are three signal components in foregoing expression, wherein the first signal component is a subchannel frequency-domain signal which is used by the receiver 300 for resolving the information transmitted by the transmitter, the second signal component is a subchannel interference signal, and the third signal component is the AWGN. Additionally, F_l(k) is defined as a FFT of the time variation of the l^th channel tap, and F_l(k) is expressed as:

$F_l\left(k\right)=\sum _{u=0}^{N-1}{h}_{r,t,l}^{\left(u\right)}{e}^{-j\frac{2\pi \mathrm{ku}}{N_c}}0\le l\le N_G\&0\le k\le N-1$
Accordingly, the subchannel interference coefficient H_r,t,k^d in the subchannel interference signal is expressed as:

$H_{r,t,k}^d=\frac{1}{N}\sum _{l=0}^{N_G}{F}_l\left(d\right)e^{-j\frac{2\pi l\left(k-d\right)}{N}}0\le k,d\le N_c-1,$
and the subchannel interference coefficient H_r,t,k⁰ is expressed as:

$\begin{array}{c}{H}_{r,t,k}^0=\frac{1}{N}\sum _{l=0}^{N_G}\sum _{u=0}^{N-1}{h}_{r,t,l}^{\left(u\right)}{e}^{-j\frac{2\pi \mathrm{lk}}{N}}\\ =\sum _{l=0}^{N_G}{h}_{r,t,l}^{\mathrm{ave}}{e}^{-j\frac{2\pi \mathrm{lk}}{N}}\end{array}$
wherein h_r,t,l^ave represents an average value of the l^th channel tap in the OFDM symbol during a valid time period.

The PPIC 340 receives a plurality of subchannel frequency-domain reception signals output by the FFTs 381˜38R and reconstructs each subchannel interference signal according to a plurality of subchannel frequency-domain estimation signals output by the MIMO detector 350 or the channel decoder 360. To be specific, the channel estimator 330 estimates that a plurality of subchannel impulse responses produced by the wireless channel can generate a plurality of coefficients for reconstructing the subchannel interference signals. The PPIC 340 receives these coefficients and reconstructs each subchannel interference signal according to the subchannel frequency-domain estimation signals and the coefficients. Next, the PPIC 340 subtracts the corresponding subchannel interference signals from the received subchannel frequency-domain reception signals to obtain the subchannel frequency-domain signals. After that, the PPIC 340 sends the subchannel frequency-domain signals to the MIMO detector 350.

The multiplexer 320 decides whether to let the subchannel frequency-domain estimation signals output by the MIMO detector 350 or the channel decoder 360 to enter the PPIC 340. It should be noted that the multiplexer 320 is not an essential element of the receiver 300, and those having ordinary knowledge in the art should understand that the multiplexer 320 may be removed, and the output terminal of the MIMO detector 350 or the channel decoder 360 for outputting the subchannel frequency-domain estimation signals may be directly coupled to the PPIC 340 after the multiplexer 320 is removed.

The MIMO detector 350 receives the subchannel frequency-domain signals output by the PPIC 340 and obtains a plurality of subchannel frequency-domain estimation signals according to the subchannel frequency-domain signals, wherein the subchannel frequency-domain estimation signals are sent back to the PPIC 340 so that the PPIC 340 can reconstruct the subchannel interference signals according to the subchannel frequency-domain estimation signals. The channel decoder 360 receives a plurality of bits that are generated by the MIMO detector 350 according to a plurality of subchannel frequency-domain signals. The channel decoder 360 decodes these bits to generate a plurality of decoded bits for a backend circuit of the receiver 300. In addition, the channel decoder 360 further generates a plurality of subchannel frequency-domain estimation signals for the PPIC 340 according to the decoded bits.

It can be understood from foregoing description that the PPIC 340, the MIMO detector 350, and the channel decoder 360 can perform multiple iterations. Through these iterations, the bits carried by the subchannel frequency-domain signals are made almost equivalent to the bits transmitted by the transmitter. Below, the iterations performed by the PPIC 340, the MIMO detector 350, and the channel decoder 360 will be described in detail.

FIG. 3 is a system block diagram illustrating part of the circuit of the PPIC 340 according to an exemplary embodiment of the present disclosure. Referring to FIG. 3, the PPIC 340 includes a plurality of subchannel interference reconstruction units and a plurality of subchannel interference cancellation units, such as the subchannel interference reconstruction units 420˜422 and the subchannel interference cancellation units 410˜412 corresponding to the (k−1)^th˜(k+1)^th subchannels (assuming that (k−1) is greater than or equal to 0 and (k+1) is smaller than N) of the r^th receiving antenna, as shown in FIG. 3. The subchannel interference reconstruction units 420˜422 are respectively coupled to the subchannel interference cancellation units 410˜412. Even though only part of the circuit of the PPIC 340 is illustrated in FIG. 3, the entire circuit of the PPIC 340 can be deduced according to FIG. 3. Similarly, even though only the iterations performed by the subchannel interference reconstruction units 420˜422 and the subchannel interference cancellation units 410˜412 are described, the iterations performed by the other subchannel interference reconstruction units and subchannel interference cancellation units can be deduced accordingly.

During the first iteration, because the MIMO detector 350 and the channel decoder 360 have not estimated the bit probabilities to generate the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽⁰⁾, Ŝ_t,k⁽⁰⁾, and Ŝ_t,k+1⁽⁰⁾ yet, the subchannel interference reconstruction units 420˜422 respectively consider the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽⁰⁾, Ŝ_t,k⁽⁰⁾, and Ŝ_t,k+1⁽⁰⁾ as 0, wherein the subchannel frequency-domain estimation signal Ŝ_t,k⁽⁰⁾ represents a predetermined k^th subchannel frequency-domain estimation signal corresponding to the t^th transmitting antenna, and the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽⁰⁾ and Ŝ_t,k+1⁽⁰⁾ can be deduced accordingly. During the first iteration, the subchannel interference reconstruction units 420˜422 directly preset the subchannel interference reconstruction signals S_ICI,r,k−1⁽¹⁾, S_ICI,r,k⁽¹⁾, and S_ICI,r,k+1⁽¹⁾ to 0, wherein the subchannel interference reconstruction signal S_ICI,r,k⁽¹⁾ represents the k^th subchannel interference reconstruction signal corresponding to the r^th receiving antenna generated during the first iteration.

The subchannel interference cancellation units 410˜412 respectively receive the subchannel frequency-domain reception signals Y_r,k−1, Y_r,k, and Y_r,k+1. During the first iteration, because the subchannel interference reconstruction signals S_ICI,r,k−1⁽¹⁾, S_ICI,r,k⁽¹⁾, and S_ICI,r,k+1⁽¹⁾ are preset to 0, the subchannel frequency-domain signals Y′_r,k−1⁽¹⁾, Y′_r,k⁽¹⁾, and Y′_r,k+1⁽¹⁾ output by the subchannel interference cancellation units 410˜412 are equivalent to the subchannel frequency-domain reception signals Y_r,k−1, Y_r,k, and Y_r,k+1, wherein the subchannel frequency-domain signal Y′_r,k−⁽¹⁾ represents the k^th subchannel frequency-domain signal corresponding to the r^th receiving antenna generated during the first iteration.

During the first iteration, the MIMO detector 350 or the channel decoder 360 estimates the bit probabilities according to the subchannel frequency-domain signals Y′_1,k⁽¹⁾, Y′_2,k⁽¹⁾, . . . , and Y′_R,k⁽¹⁾ to generate the subchannel frequency-domain estimation signal Ŝ_t,k⁽¹⁾. Similarly, the MIMO detector 350 or the channel decoder 360 estimates the bit probabilities to generate the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽¹⁾ and Ŝ_t,k+1⁽¹⁾.

Next, the subchannel interference reconstruction unit 421 generates the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k−i+1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ during the i^th iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal Ŝ_t,q^(x) represents the q^th subchannel frequency-domain estimation signal corresponding to the t^th transmitting antenna generated during the x^th iteration, x is an integer between 1 and (i−1), q=k−i+x or q=k+i−x, and the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ represents the k^th subchannel interference reconstruction signal corresponding to the r^th receiving antenna generated during the i^th iteration. Besides, the i^th iteration performed by the subchannel interference reconstruction units 420 and 422 can be deduced accordingly.

To be specific, the method for generating the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ can be referred to the expression of the subchannel interference signal in aforementioned expression of the subchannel frequency-domain reception signal Y_r,k, and accordingly the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ can be expressed as:
S_ICI,r,k⁽ⁱ⁾=H_r,t,k¹·Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾+H_r,t,k²·Ŝ_t,((k−2))N⁽ⁱ⁻²⁾+ . . . +H_r,t,k^N−2·Ŝ_t,((k−2))N⁽ⁱ⁻²⁾+H_r,t,k^N−1·Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾
If it is assumed that the subchannel impulse responses estimated by the channel estimator 330 are perfect, the coefficients H_r,t,k¹˜H_r,t,k^N−1 generated by the channel estimator 330 also have ideal values. Thus, the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ is close to the k^th practical subchannel interference reconstruction signal corresponding to the r^th receiving antenna, and along with the increase of the value of i, the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ will get closer to the k^th practical subchannel interference reconstruction signal corresponding to the r^th receiving antenna.

During the i^th iteration, the subchannel interference cancellation unit 411 obtains the subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ according to the subchannel frequency-domain reception signal Y_r,k and the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾. To be specific, the subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ is equal to the subchannel frequency-domain reception signal Y_r,k minus the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾, wherein the subchannel frequency-domain signal Y_r,k⁽ⁱ⁾ represents the k^th subchannel frequency-domain signal corresponding to the r^th receiving antenna generated during the i^th iteration. In addition, the i^th iteration performed by the subchannel interference cancellation units 410 and 412 can be deduced accordingly.

It should be mentioned that in the exemplary embodiment illustrated in FIG. 3, the subchannel interference reconstruction units 420˜422 respectively have a memory. Taking the subchannel interference reconstruction unit 421 as an example, during the i^th iteration, the subchannel interference reconstruction unit 420 receives and stores the subchannel frequency-domain estimation signal Ŝ_t,k⁽ⁱ⁻¹⁾. During the i^th iteration, the subchannel interference reconstruction unit 421 transmits the subchannel frequency-domain estimation signal Ŝ_t,k⁽ⁱ⁻¹⁾ currently stored in the memory to the subchannel interference reconstruction units 420 and 422.

Besides, during the i^th iteration, the subchannel interference reconstruction unit 421 transmits the subchannel frequency-domain estimation signals Ŝ_{t,((k−i+2))N}⁽¹⁾, Ŝ_{t,((k−i+3))N}⁽²⁾, . . . , and Ŝ_t,((k−1))N⁽ⁱ⁻²⁾ that are stored in the memory during the previous (i−2) iterations to the subchannel interference reconstruction unit 422, and the subchannel interference reconstruction unit 421 transmits the subchannel frequency-domain estimation signals Ŝ_{t,((k−i+2))N}⁽¹⁾, Ŝ_{t,((k+i−3))N}⁽²⁾, . . . , and Ŝ_t,((k+1))N⁽ⁱ⁻²⁾ that are stored in the memory during the previous (i−2) iterations to the subchannel interference reconstruction unit 420.

In addition, during the i^th iteration, the subchannel interference reconstruction unit 421 receives the subchannel frequency-domain estimation signals Ŝ_{t,((k−i+1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾ from the subchannel interference reconstruction unit 420 and stores the subchannel frequency-domain estimation signals Ŝ_{t,((k−i+1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾ into the memory. Similarly, the subchannel interference reconstruction unit 421 receives the subchannel frequency-domain estimation signals Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ from the subchannel interference reconstruction unit 422 and stores the subchannel frequency-domain estimation signals Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ into the memory.

FIGS. 4A˜4C are diagrams respectively illustrating how the subchannel interference cancellation units 410˜412 and the subchannel interference reconstruction units 420˜422 perform the first to the third iterations. Referring to FIG. 4A, during the first iteration, because the MIMO detector 350 and the channel decoder 360 have not estimated the bit probabilities to generate the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽⁰⁾, Ŝ_t,k⁽⁰⁾, and Ŝ_t,k+1⁽⁰⁾, the subchannel interference reconstruction units 420˜422 respectively consider the subchannel frequency-domain estimation signals Ŝ_t,k−1⁽⁰⁾, Ŝ_t,k⁽⁰⁾, and Ŝ_t,k+1⁽⁰⁾ as 0 and directly preset the subchannel interference reconstruction signals S_ICI,r,k−1⁽¹⁾, S_ICI,r,k⁽¹⁾, and S_ICI,r,k+1⁽¹⁾ to 0. The subchannel interference cancellation units 410˜412 respectively receive the subchannel frequency-domain reception signals Y_r,k−1, Y_r,k, and Y_r,k+1 and outputs the subchannel frequency-domain signals Y′_r,k−1⁽¹⁾, Y′_r,k⁽¹⁾, and Y′_r,k+1⁽¹⁾, wherein the subchannel frequency-domain signals Y′_r,k−1⁽¹⁾, Y′_r,k⁽¹⁾, and Y′_r,k+1⁽¹⁾ are used for replacing the subchannel frequency-domain reception signals Y_r,k−1, Y_r,k, and Y_r,k+1.

It should be noted that the MIMO detector 350 works along with the PPIC 340 to perform the iterations. Thus, when the PPIC 340 performs the first iteration, the channel estimator 330 needs only to calculate the coefficients H,r,t,0⁰˜H,r,t,1023⁰ (assuming N is 1024), namely, to generate the following subchannel interference coefficient matrix:

$\mathrm{Subcarrier}\mathrm{Index}k\left[\begin{array}{ccccccc}\stackrel{\underset{\underset{\mathrm{Term}}{\mathrm{Signal}}}{d=0}}{H_{r,t,0}^0}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{1}^{\mathrm{st}}}}{d=1}}{n.a.}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{2}^{\mathrm{nd}}}}{d=2}}{n.a.}& \dots & \dots & \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{2}^{\mathrm{nd}}}}{d=1022}}{n.a.}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{1}^{\mathrm{st}}}}{d=1023}}{n.a.}\\ H_{r,t,1}^0& n.a.& n.a.& \dots & \dots & n.a.& n.a.\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ H_{r,t,1022}^0& n.a.& n.a.& \dots & \dots & n.a.& n.a.\\ H_{r,t,1023}^0& n.a.& n.a.& \dots & \dots & n.a.& n.a.\end{array}\right],$
wherein n.a. means that the corresponding coefficient is not to be generated.

Next, referring to FIG. 4B, during the second iteration, the subchannel interference reconstruction unit 421 generates the subchannel interference reconstruction signal S_ICI,r,k⁽²⁾ according to two subchannel frequency-domain estimation signals Ŝ_t,k−1⁽¹⁾ and Ŝ_t,k+1⁽¹⁾. The subchannel interference cancellation unit 411 obtains the subchannel frequency-domain signal Y′_r,k⁽²⁾ according to the subchannel frequency-domain reception signal Y_r,k and the subchannel interference reconstruction signal S_ICI,r,k⁽²⁾. Similarly, the subchannel interference reconstruction unit 420 generates the subchannel interference reconstruction signal S_ICI,r,k−1⁽²⁾ according to two subchannel frequency-domain estimation signals Ŝ_t,k−2⁽¹⁾ and Ŝ_t,k⁽¹⁾ (assuming that (k−2) is greater than or equal to 0). The subchannel interference cancellation unit 410 obtains the subchannel frequency-domain signal Y′_r,k−1⁽²⁾ according to the subchannel frequency-domain reception signal Y_r,k−1 and the subchannel interference reconstruction signal S_ICI,r,k−1⁽²⁾. The subchannel interference reconstruction unit 422 generates the subchannel interference reconstruction signal S_ICI,r,k+1⁽²⁾ according to two subchannel frequency-domain estimation signals Ŝ_t,k⁽¹⁾ and Ŝ_t,k+2⁽¹⁾ (assuming that (k+2) is smaller than N). The subchannel interference cancellation unit 412 obtains the subchannel frequency-domain signal Y′_r,k+1⁽²⁾ according to the subchannel frequency-domain reception signal Y_r,k+1 and the subchannel interference reconstruction signal S_ICI,r,k+1⁽²⁾.

During the second iteration, the channel estimator 330 needs only to calculate the coefficients H_r,t,0⁰˜H_r,t,1023⁰, H_r,t,0¹˜H_r,t,1023¹, and H_r,t,0¹⁰²³˜H_r,t,1023¹⁰²³ (assuming N is 1024), namely, to generate the following subchannel interference coefficient matrix:

$\mathrm{Subcarrier}\mathrm{Index}k\left[\begin{array}{ccccccc}\stackrel{\underset{\underset{\mathrm{Term}}{\mathrm{Signal}}}{d=0}}{H_{r,t,0}^0}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{1}^{\mathrm{st}}}}{d=1}}{H_{r,t,0}^1}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{2}^{\mathrm{nd}}}}{d=2}}{n.a.}& \dots & \dots & \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{2}^{\mathrm{nd}}}}{d=1022}}{n.a.}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{1}^{\mathrm{st}}}}{d=1023}}{H_{r,t,0}^{1023}}\\ H_{r,t,1}^0& H_{r,t,1}^1& n.a.& \dots & \dots & n.a.& H_{r,t,1}^{1023}\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ H_{r,t,1022}^0& H_{r,t,1022}^1& n.a.& \dots & \dots & n.a.& H_{r,t,1022}^{1023}\\ H_{r,t,1023}^0& H_{r,t,1023}^1& n.a.& \dots & \dots & n.a.& H_{r,t,1023}^{1023}\end{array}\right]$

Next, referring to FIG. 4C, during the third iteration, the subchannel interference reconstruction unit 421 generates the subchannel interference reconstruction signal S_ICI,r,k⁽³⁾ according to four subchannel frequency-domain estimation signals Ŝ_t,k−2⁽¹⁾, Ŝ_t,k−1⁽²⁾, Ŝ_t,k+1⁽²⁾, and Ŝ_t,k−2⁽¹⁾. The subchannel interference cancellation unit 411 obtains the subchannel frequency-domain signal Y′_r,k⁽³⁾ according to the subchannel frequency-domain reception signal Y_r,k and the subchannel interference reconstruction signal S_ICI,r,k⁽³⁾. Similarly, the subchannel interference reconstruction unit 420 generates the subchannel interference reconstruction signal S_ICI,r,k−1 according to four subchannel frequency-domain estimation signals Ŝ_t,k−3⁽¹⁾, Ŝ_t,k−2⁽²⁾, Ŝ_t,k⁽²⁾, and Ŝ_t,k+1⁽¹⁾ (assuming that (k−3) is greater than or equal to 0). The subchannel interference cancellation unit 410 obtains the subchannel frequency-domain signal Y′_r,k−1⁽³⁾ according to the subchannel frequency-domain reception signal Y_r,k−1 and the subchannel interference reconstruction signal S_ICI,r,k−1⁽³⁾. The subchannel interference reconstruction unit 422 generates the subchannel interference reconstruction signal S_ICI,r,k+1⁽³⁾ according to four subchannel frequency-domain estimation signals Ŝ_t,k−1⁽¹⁾, Ŝ_t,k⁽²⁾, Ŝ_t,k+2⁽²⁾, and Ŝ_t,k+3⁽¹⁾ (assuming that (k+3) is smaller than N). The subchannel interference cancellation unit 412 obtains the subchannel frequency-domain signal Y′_r,k+1⁽³⁾ according to the subchannel frequency-domain reception signal Y_r,k+1 and the subchannel interference reconstruction signal S_ICI,r,k+1⁽³⁾.

During the third iteration, the channel estimator 330 needs only to calculate the coefficients H_r,t,0⁰˜H_r,t,1023⁰, H_r,t,0¹˜H_r,t,1023¹, H_r,t,0²˜H_r,t,1023², H_r,t,0¹⁰²²˜H_r,t,1023¹⁰²², and H_r,t,0¹⁰²³˜H_r,t,1023¹⁰²³ (assuming that N is 1024), namely, to generate the following subchannel interference coefficient matrix:

$\mathrm{Subcarrier}\mathrm{Index}k\hspace{1em}\left[\begin{array}{ccccccc}\stackrel{\underset{\underset{\mathrm{Term}}{\mathrm{Signal}}}{d=0}}{H_{r,t,0}^0}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{1}^{\mathrm{st}}}}{d=1}}{H_{r,t,0}^1}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Left}{2}^{\mathrm{nd}}}}{d=2}}{H_{r,t,0}^2}& n.a.& n.a.& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{2}^{\mathrm{nd}}}}{d=1022}}{H_{r,t,0}^{1022}}& \stackrel{\underset{\underset{\mathrm{ICI}\mathrm{Term}}{\mathrm{Right}{1}^{\mathrm{st}}}}{d=1023}}{H_{r,t,0}^{1023}}\\ H_{r,t,1}^0& H_{r,t,1}^1& H_{r,t,1}^2& n.a.& n.a.& H_{r,t,1}^{1022}& H_{r,t,1}^{1023}\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ ⋮& ⋮& ⋮& \dots & \dots & ⋮& ⋮\\ H_{r,t,1022}^0& H_{r,t,1022}^1& H_{r,t,1022}^2& n.a.& n.a.& H_{r,t,1022}^{1022}& H_{r,t,1022}^{1023}\\ H_{r,t,1023}^0& H_{r,t,1023}^1& H_{r,t,1023}^2& n.a.& n.a.& H_{r,t,1023}^{1022}& H_{r,t,1023}^{1023}\end{array}\right]$
It can be understood from foregoing example that the PPIC 340 provided by exemplary embodiments of the present disclosure cancels interference from two more adjacent subchannels during each iteration. Besides, the channel estimator 330 only calculates two more columns of coefficients during each iteration than a previous iteration.

The PPIC 340 can work together with any type of MIMO detector 350 to completely filter out subchannel interferences. However, an implementation of the MIMO detector 350 and the iterations performed by the MIMO detector 350 will be described with reference to another exemplary embodiment.

The MIMO detector 350 determines a plurality of bits b_t,k,1, b_t,k,2, . . . , and b_t,k,M of the k^th subchannel of the t^th transmitting antenna according to a plurality of k^th subchannel frequency-domain signals Y′_1,k^(y), Y′_2,k^(y), and Y′_R,k^(y) of R receiving antennas when the PPIC 340 performs the y^th iteration (y is an integer greater than or equal to 1), wherein the bits b_t,k,1, b_t,k,2, . . . , and b_t,k,M of the k^th subchannel of the t^th transmitting antenna are used for determining the subchannel frequency-domain estimation signal Ŝ_t,k^(y), and the subchannel frequency-domain estimation signal Ŝ_t,k^(y) can be determined by the MIMO detector 350 or the channel decoder 360.

In the present exemplary embodiment, the MIMO detector 350 may be a MIMO detector using a message passing algorithm (MPA). The messages to be passed through the MPA are probability messages, wherein the probability messages are passed through each iteration so that the performance of the receiver 300 adopting the MIMO detector 350 can be improved.

Herein a bit vector b is defined as a set of bits from the first bit of the k^th subchannel of the first transmitting antenna to the last bit of the k^th subchannel of the T^th transmitting antenna (i.e., b={b₀˜b_MT−1}, and for the convenience of description, the subscript k is omitted). When the k^th subchannel frequency-domain signal corresponding to the r^th receiving antenna output by the PPIC 340 is Y′_r,k^(y), a probability message is generated for the bit b_n according to the subchannel frequency-domain signal Y′_r,k^(y), wherein the bit b_n represents the n^th bit in the bit vector b. Herein the probability message generated for the bit b_n by the subchannel frequency-domain signal Y′_r,k^(y) is defined as η_(r→n),k(bn). Namely, the probability message η_(r→n),k(bn) represents the probability of that the bit b_n is 0 or 1 when the subchannel frequency-domain reception signal is Y_r,k. The probability message η_(r→n),k(b_n) can be expressed as:

${\eta }_{\left(r⟶n\right),k}\left(b_n\right)=P\lfloor b_n=b❘Y_{r,k}^{\prime \left(y\right)}\rfloor ,b=0\mathrm{or}1\propto \sum _{{\omega }_0^{\mathrm{MT}-1\backslash n}\in \Omega ,\rho \in A_{r,k}}p\left(Y_{r,k}^{\prime \left(y\right)}❘{\phi }_{r,k}=\rho \right)\prod _{j=0,j\ne n}^{\mathrm{MT}-1}{\mu }_{\left(j⟶r\right),k}\left(b_j\right)=\sum _{{\omega }_0^{\mathrm{MT}-1\backslash n}\in \Omega ,\rho \in A_{r,k}}p\left(Y_{r,k}^{\prime \left(y\right)}❘{\phi }_{r,k}=\rho \right)\left(\mathrm{for}\mathrm{the}{1}^{\mathrm{st}}\mathrm{iteration}\right)\approx \mathrm{MAX}\left[\sum _{{\omega }_0^{\mathrm{MT}-1\mathrm{\backslash n}}\in \Omega ,\rho \in A_{r,k}}p\left(Y_{r,k}^{\prime \left(y\right)}❘{\phi }_{r,k}=\rho \right)·\prod _{j=0,j\ne n}^{\mathrm{MT}-1}{\mu }_{\left(j⟶r\right),k}\left(b_j\right)\right]\left(\mathrm{for}\mathrm{Max}\mathrm{Product}\mathrm{Rule}\right)$
wherein ρ belongs to one of the set A_r,k, ω₀^MT−1\n belongs to one of the set Ω. φ_r,k is an MIMO channel output alphabet of the k^th subchannel of the r^th receiving antenna (i.e., φ_r,k=Σ_t=1^Th_tS_t,k), h_t is a subchannel impulse response matrix, A_r,k represents all the possible sets of MIMO channel output alphabets of the k^th subchannel of the r^th receiving antenna, ω₀^MT−1\n represents the bit input vectors in the MIMO channel besides ω_n (i.e., ω₀^MT−1\n={ω₀, . . . , ω_n−1, ω_n+1, . . . , ω_MT−1}), and Ω is the set of all the bit input vectors in the MIMO channel. The probability message μ_(j→r),k(b_j) represents the probability message generated by the bit b_j for the subchannel frequency-domain reception signal Y_r,k, and which is equivalent to the probability of that the bit b_j is 1 or 0.

When the MIMO detector 350 performs the first iteration, the MIMO detector 350 uses every probability p(Y′_r,k^(1)=ρ) to obtain the probability message η_(r→n),k(b_n). It should be noted that every time after the PPIC 340 performs an iteration, the MIMO detector 350 performs several sub iterations to calculate the possible probability of each bit and determine the bits to be output. Thus, the MIMO detector 350 or the channel decoder 360 generates a plurality of subchannel frequency-domain estimation signals according to these bits to be used by the PPIC 340 during each iteration.

During the sub iterations after the first sub iteration, the MIMO detector 350 has to consider the probability message μ_(j→r),k(b_j) to obtain the probability message η_(r→n),t,k(b). In other words, during the sub iterations after the first sub iteration, the MIMO detector 350 has to consider each probability p(Y′_r,k^(y)|φ_r,k=ρ) and each probability message μ_(j→r)k(b_j) generated during the previous sub iteration to obtain the probability message η_(r→n),k(b_n).

The probability message μ_(n→p),k(b_n) represents a probability message generated by the bit b_n for the subchannel frequency-domain reception signal Y_p,k and which is equivalent to the probability of that the bit b_n is 1 or 0. The probability message μ_(n→p),k(b_n) is expressed as:

$\begin{array}{c}{\mu }_{\left(n⟶p\right),k}\left(b_n\right)=P\left(b_n=b\right)\\ =\frac{P_a\left[b_n=b\right]·\prod _{r=0,r\ne p}^{R-1}P\left[b_n=b❘Y_{r,k}^{\prime \left(y\right)}\right]}{\begin{array}{c}{P}_a\left[b_n=0\right]·\prod _{r=0,r\ne p}^{R-1}P\left[b_n=0❘Y_{r,k}^{\prime \left(y\right)}\right]+\\ P_a\left[b_n=1\right]·\prod _{r=0,r\ne p}^{R-1}P\left[b_n=1❘Y_{r,k}^{\prime \left(y\right)}\right]\end{array}}\\ =\frac{{\eta }_n\left(b\right)·\prod _{r=0,r\ne p}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(b\right)}{\begin{array}{c}{\eta }_n\left(0\right)·\prod _{r=0,r\ne p}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(0\right)+{\eta }_n\left(1\right)·\\ \prod _{r=0,r\ne p}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(1\right)\end{array}}\end{array}$
wherein η_n(b) is an a-priori probability of the bit b_n, and which is equal to P_a[b_n=b]. During the first sub iteration, the a-priori probability η_n(b) is initialized to 0.5. During the sub iterations of the w^th iteration (w is an integer greater than 1), the a-priori probability η_n(b) is a result obtained by the channel decoder 360, and the a-priori probability η_n(b) is set to 0.5 if the channel decoder 360 does not offer any result.

FIG. 5 is a diagram of a MPA according to an exemplary embodiment of the present disclosure, wherein a MIMO communication system with four receiving antennas is described as an example. Referring to FIG. 5, the function node 501 represents a function of receiving the probability p(Y′_4,k^(y)|φ_4,k=ρ) calculated according to the subchannel frequency-domain reception signal Y′_r,k^(y) and the probability message η_(4→1),k(b₁) calculated according to the probability messages μ_(2→4),k(b₂), μ_(3→4),k(b₃), and μ_(4→4),k(b₄) and transmitting the probability message η_(4→1),k(b₁) to the bit node 500. Then, the bit node 500 calculates the probability message μ_(1→1)k(b₁) according to the a-priori probability η₁(b₁) and the probability messages η_(2→1),k(b₁), η_(3→1),k(b₁), and η_(4→1),k(b₁) and transmits the probability message μ_(1→1),k(b₁) to the function node (not shown) for receiving the subchannel frequency-domain reception signal Y_1,k.

In short, the MIMO detector 350 first calculates a probability message η_(r→n),k(b_n) during each sub iteration and then calculates a probability message μ_(nΘp),k(b_n). During the next sub iteration, the MIMO detector 350 calculates the probability message η_(r→n)k(b_n) according to the probability message μ_(nΘp),k(b_n) calculated during the previous sub iteration and calculates the probability message μ_(nΘp),k(b_n) according to the probability message η_(r→n),k(b_n). After performing certain number of sub iterations, the MIMO detector 350 calculates the final probability μ_n,k(b_n) of that the bit b_n is 1 or 0. The final probability μ_n,k(b_n) can be expressed as:

${\mu }_{n,k}\left(b\right)=P\left(b_n=b\right)=\frac{{\eta }_n\left(b\right)·\prod _{r=0}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(b\right)}{{\eta }_n\left(0\right)·\prod _{r=0}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(0\right)+{\eta }_n\left(1\right)·\prod _{r=0}^{R-1}{\eta }_{\left(r⟶n\right),k}\left(1\right)}$
After calculating the final probability μ_n,k(b_n) of the bit b_n, the MIMO detector 350 makes a hard decision or a soft decision to the bit b_n according to the final probability μ_n,k(b_n) of the bit b_n to determine whether the bit b_n is 0 or 1.

If enough number of iterations has been performed or the bit error rate (BER) has met the requirement, the PPIC 340 does not perform any iteration. In this case, the MIMO detector 350 makes a hard decision to the bit b_n according to the final probability μ_n,k(b_n) of the bit b_n. Contrarily, if the PPIC 340 continues to perform iterations, the MIMO detector 350 makes a soft decision to the bit b_n according to the final probability μ_n,k(b_n) of the bit b_n.

When the MIMO detector 350 makes the soft decision to the bit b_n according to the final probability μ_n,k(b_n) of the bit b_n, the bit b_n is determined through following expression:

${\hat{b}}_n=\tanh \left\{0.5·\ln \left[\frac{{\mu }_{n,k}\left(1\right)}{{\mu }_{n,k}\left(0\right)}\right]\right\}$
wherein {circumflex over (b)}_n represents the output value of the bit b_n. When the MIMO detector 350 makes the hard decision to the bit b_n according to the final probability μ_n,k(b_n) of the bit b_n, the bit b_n is determined through following expression:

${\tilde{b}}_n=\{\begin{array}{cc}1,& {\mu }_{n,k}\left(1\right)\ge {\mu }_{n,k}\left(0\right)\\ 0,& {\mu }_{n,k}\left(1\right)<{\mu }_{n,k}\left(0\right)\end{array}$
wherein {tilde over (b)}_n represents the output value of the bit b_n.

The MIMO detector 350 generates a plurality of subchannel frequency-domain estimation signals according to a plurality of bits generated when it makes the soft decision, wherein these subchannel frequency-domain estimation signals allow the PPIC 340 to reconstruct the subchannel interference signals. If the modulation order is 2, a quadrature phase shift keying (QPSK) modulation technique can be adopted. Herein the subchannel frequency-domain estimation signal Ŝ_t,k⁽ⁱ⁾ generated by the MIMO detector 350 can be expressed as:

${\hat{S}}_{t,k}^{\left(i\right)}=\frac{1}{\sqrt{2}}·{\hat{b}}_{M\left(t-1\right)}+j\frac{1}{\sqrt{2}}·{\hat{b}}_{M\left(t-1\right)+1}$
If the modulation order is 4, a 16 quadrature amplitude modulation (16-QAM) can be adopted. Herein the subchannel frequency-domain estimation signal Ŝ_t,k⁽ⁱ⁾ generated by the MIMO detector 350 can be expressed as:

${\hat{S}}_{t,k}^{\left(i\right)}=-\frac{1}{\sqrt{10}}·{\hat{b}}_{M\left(t-1\right)}·\left(2+{\hat{b}}_{M\left(t-1\right)+2}\right)-j\frac{1}{\sqrt{10}}·{\hat{b}}_{M\left(t-1\right)+1}·\left(2+{\hat{b}}_{M\left(t-1\right)+3}\right)$

In addition, it should be noted that the MIMO detector 350 working along with the PPIC 340 may not be a MIMO detector using MPA. Instead, the PPIC 340 may also work along with different types of MIMO detectors. Similarly, the MIMO detector 350 using the MPA may also work along with a different type of PPIC.

Below, a progressive parallel interference cancellation method provided by the present disclosure will be described with reference to an exemplary embodiment of the present disclosure. FIG. 6 is a flowchart of a progressive parallel interference cancellation method according to an exemplary embodiment of the present disclosure. The steps in FIG. 6 can be accomplished by the PPIC 340 and the MIMO detector 350 illustrated in FIG. 2.

First, in step S600, a subchannel frequency-domain estimation signal Ŝ_t,k⁽⁰⁾ is initialized to 0. Then, in step S601, a subchannel interference signal S_ICI,r,k⁽⁰⁾ or S_ICI,r,k⁽ⁱ⁾ is reconstructed, wherein during the first iteration, the subchannel interference signal S_ICI,r,k⁽⁰⁾ is preset to 0, and during the i^th iteration, a subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ is generated according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k−i+1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾, wherein the method of generating the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ according to the (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k−i+1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ can be referred to foregoing description and will not be described herein.

Next, in step S602, the subchannel interference signal component in the received subchannel frequency-domain reception signal Y_r,k is cancelled, so as to generate the subchannel frequency-domain signal and perform a MIMO detection. During the first iteration, the subchannel frequency-domain signal Y′_r,k⁽¹⁾ is equal to the subchannel frequency-domain reception signal Y_r,k, and during the i^th iteration, the subchannel frequency-domain signal Y′_r,k⁽ⁱ⁾ is equal to the subchannel frequency-domain reception signal Y_r,k minus the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾.

Thereafter, in step S603, the MIMO detection is performed according to a plurality of subchannel frequency-domain signals, so as to calculate the probability of each bit. For example, during the y^th iteration, the MIMO detection is performed to the subchannel frequency-domain signals Y′_1,k^(y), Y′_2,k^(y), . . . , and Y′_R,k^(y) to calculate the probabilities of the first bit in the k^th subchannel of the first transmitting antenna to the last bit in the k^th subchannel of the R^th transmitting antenna. The probability of each bit may be calculated through aforementioned MPA or other methods.

Next, in step S604, whether the iteration is still to be performed is determined. If no more iteration is to be performed, step S608 is executed; otherwise, step S605 is executed. Whether to perform more iterations may be determined by determining whether the number of iterations satisfies a predetermined iteration number or whether a SNR or BER calculated by a backend circuit satisfies the requirement of the entire communication system.

If the iteration is still to be performed, in step S605, a soft decision is made according to the probability of each bit so as to determine an output value of each bit. The method for determining the output value of each bit has been described above therefore will not be described herein. Thereafter, in step S606, the bits generated through the soft decision are mapped to generate a plurality of subchannel frequency-domain estimation signals, wherein how to map the bits to generate the subchannel frequency-domain estimation signals has been described above therefore will not be described herein.

Next, in step S607, the subchannel frequency-domain estimation signals generated during previous iterations are passed so that in step S601, the subchannel interference reconstruction signal S_ICI,r,k⁽ⁱ⁾ can be generated simply according to (2×i−2) subchannel frequency-domain estimation signals Ŝ_{t,((k+i−1))N}⁽¹⁾, Ŝ_{t,((k−i+2))N}⁽²⁾, . . . , Ŝ_t,((k−2))N⁽ⁱ⁻²⁾, Ŝ_t,((k−1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+1))N⁽ⁱ⁻¹⁾, Ŝ_t,((k+2))N⁽ⁱ⁻²⁾, . . . , Ŝ_{t,((k+i−2))N}⁽²⁾, Ŝ_{t,((k+i−1))N}⁽¹⁾ during the i^th iteration. The method for passing the subchannel frequency-domain estimation signals generated during previous iterations have been described above therefore will not be described herein.

The subchannel interference signal component is greatly reduced after several iterations are performed. Then, when the iteration is not to be performed anymore, in step S608, a hard decision is performed according to the final probability of each bit generated in the step S603, so as to generate the output value of each bit.

As described above, the present disclosure provides a PPIC and a method and a receiver thereof, wherein each subchannel interference reconstruction signal is reconstructed through iterations, and a corresponding subchannel interference reconstruction signal is subtracted from each subchannel frequency-domain reception signal to obtain a subchannel frequency-domain signal. According to the present disclosure, subchannel interferences can be cancelled without re-performing channel coding or estimating the SNR or frequency offset. In addition, the technique provided by the present disclosure has reduced calculation complexity therefore is adaptable to very large scale integrated circuits (VLSI).

It will be apparent to those skilled in the art that various modifications and variations can be made to the structure of the present disclosure without departing from the scope or spirit of the disclosure. In view of the foregoing, it is intended that the present disclosure cover modifications and variations of this disclosure provided they fall within the scope of the following claims and their equivalents.

Claims

1. A progressive parallel interference canceller (PPIC), adaptable to a receiver in a multiple input multiple output (MIMO) communication system based on an orthogonal frequency-division multiplexing (OFDM) modulation technique, wherein the receiver performs a plurality of iterations, the PPIC comprising:

a plurality of subchannel interference reconstruction units, wherein the kth subchannel interference reconstruction unit corresponding to a rth receiving antenna generates a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna during the ith iteration according to (2×i−2) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+1))N(i−1), Ŝt,((k+2))N(i−2),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1), wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal Ŝt,q(x) represents a qth subchannel frequency-domain estimation signal corresponding to a tth transmitting antenna generated during the xth iteration, and ((•))N represents a calculation of dividing by N to obtain a remainder, N is a number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), and q=k−i+x or q=k+i−x; and

a plurality of subchannel interference cancellation units, wherein the kth subchannel interference cancellation unit corresponding to the rth receiving antenna is coupled to the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna, and the kth subchannel interference cancellation unit obtains a kth subchannel frequency-domain signal Y′r,k(i) corresponding to the rth receiving antenna according to a kth subchannel frequency-domain reception signal Yr,k and the kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna during the ith iteration.

2. The PPIC according to claim 1, wherein during the 1st iteration, the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna presets a kth subchannel interference reconstruction signal SICI,r,k(1) corresponding to the rth receiving antenna to 0, and the kth subchannel interference cancellation unit corresponding to the rth receiving antenna obtains a kth subchannel frequency-domain signal Y′r,k(1) corresponding to the rth r,k receiving antenna according to the kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna.

3. The PPIC according to claim 2 further comprising:

a MIMO detector, coupled to the subchannel interference reconstruction units and the subchannel interference cancellation units, for determining a plurality of bits b0, b1,..., and bMT−1 corresponding to kth subchannels of the transmitting antennas according to the kth subchannel frequency-domain signals Y′1,k(i), Y′2,k(i),..., and Y′R,k(i) of the receiving antennas when the PPIC performs the ith iteration, wherein M is a bit modulation order, R is a number of the receiving antennas, T is a number of the transmitting antennas, and the bits bM(t−1), bM(t−1)+1,..., and bMt−1 corresponding to the kth subchannel of the tth transmitting antenna are used for determining the subchannel frequency-domain estimation signal Ŝt,k(i).

4. The PPIC according to claim 3, wherein the MIMO detector is a MIMO detector using a message passing algorithm (MPA).

5. The PPIC according to claim 3, wherein the MIMO detector performs a plurality of sub iterations during the ith iteration, and during each of the sub iterations, the MIMO detector generates a probability message for the bit bn according to the subchannel frequency-domain signal Y′r,k(i) and generates a probability message for the subchannel frequency-domain signal Y′p,k(i) according to the bit ba, and during the last sub iteration, the MIMO detector generates probabilities of the bits b0, b1,..., and bMT−1 according to the probability messages, wherein p is an integer between 1 and T, and a and n are integers between 0 and (MT−1).

6. The PPIC according to claim 5, wherein if the ith iteration is the last iteration, the MIMO detector makes a hard decision according to the probabilities of the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1, and if the ith iteration is not the last iteration, the MIMO detector makes a soft decision according to the probabilities of the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1.

7. The PPIC according to claim 3 further comprising a channel decoder coupled to the MIMO detector, wherein the channel decoder determines the subchannel frequency-domain estimation signal Ŝt,k(i) according to the bits bm(t−1), bM(t−1)+1,..., and bMt−1 of the kth subchannel of the tth transmitting antenna.

8. The PPIC according to claim 3, wherein the MIMO detector determines the subchannel frequency-domain estimation signal Ŝt,k(i) according to the bits bM(t−1), bM(t−1)+1,..., and bMt−1 of the kth subchannel of the tth transmitting antenna.

9. The PPIC according to claim 1, wherein a kth subchannel frequency-domain signal Y′r,k(i) corresponding to the rth receiving antenna is equivalent to a kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna minus the kth subchannel interference reconstruction signal SICI,r,k(i).

10. The PPIC according to claim 1, wherein the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna has a memory, and the kth subchannel interference reconstruction unit of the rth receiving antenna stores the subchannel frequency-domain estimation signals Ŝt,((k+i−1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+1))N(i−1), Ŝt,((k+2))N(i−2),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1) into the memory, transmits the subchannel frequency-domain estimation signals Ŝt,((k−i+2))N(1), Ŝt,((k−i+3))N(2),..., Ŝt,((k−1))N(i−2), Ŝt,((k))N(i−1) to the (k+1)th subchannel interference reconstruction unit corresponding to the rth receiving antenna, and transmits the subchannel frequency-domain estimation signals Ŝt,((k+i−2))N(1), Ŝt,((k+i−3))N(2),..., Ŝt,((k+1))N(i−2), Ŝt,((k))N(i−1) to the (k−1)th subchannel interference reconstruction unit corresponding to the rth receiving antenna.

11. The PPIC according to claim 1 further comprising a channel estimator, wherein the channel estimator estimates a plurality of channel impulse responses of a wireless channel and calculates a plurality of coefficients according to the channel impulse responses, wherein the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna generates a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna according to the (2×i−2) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k+2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+1))N(i−1),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1) and the coefficients during the ith iteration.

12. A receiver, adaptable to a MIMO communication system based on an OFDM modulation technique, wherein the receiver performs a plurality of iterations and comprises:

a plurality of receiving antennas;

a plurality of OFDM demodulators, wherein each of the OFDM demodulators is coupled to one of the receiving antennas;

a PPIC, coupled to the OFDM demodulators, the PPIC comprising: a plurality of subchannel interference reconstruction units, wherein the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna generates a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna according to (2×i−2) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k+1))N(i−1), Ŝt,((k+2))N(i−2), Ŝt,((k−1))N(i−1),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1) during the ith iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal represents the qth subchannel frequency-domain estimation signal corresponding to a tth transmitting antenna generated during the xth iteration, ((•))N represents a calculation of dividing by N to obtain a remainder, N is a number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), and q=k−i+x or q=k+i−x; and a plurality of subchannel interference cancellation units, wherein the kth subchannel interference cancellation unit corresponding to the rth receiving antenna is coupled to the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna, and the kth subchannel interference cancellation unit obtains a kth subchannel frequency-domain signal Yr,k(i) corresponding to the rth receiving antenna according to a kth subchannel frequency-domain reception signal Yr,k and the kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna during the ith iteration; and

a MIMO detector, coupled to the subchannel interference reconstruction units and the subchannel interference cancellation units, for determining a plurality of bits b0, b1,..., and bMT−1 corresponding to kth subchannels of the transmitting antennas according to the kth subchannel frequency-domain signals Y′1,k(i), Y′2,k(i),..., and Y′R,k(i) of the receiving antennas when the PPIC performs the ith iteration, wherein M is a bit modulation order, R is a number of the receiving antennas, T is a number of the transmitting antennas, and the bits bM(t−1), bM(t−1)+1,..., and bMt−1 corresponding to the kth subchannel of the tth transmitting antenna are used for determining the subchannel frequency-domain estimation signal Ŝt,k(i).

13. The receiver according to claim 12, wherein when the PPIC performs the 1st iteration, the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna presets a kth subchannel interference reconstruction signal SICI,r,k(1) corresponding to the rth receiving antenna to 0, and the kth subchannel interference cancellation unit corresponding to the rth receiving antenna obtains a kth subchannel frequency-domain signal Y′r,k(1) corresponding to the rth receiving antenna according to the kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna.

14. The receiver according to claim 12, wherein a kth subchannel frequency-domain signal Y′r,k(i) corresponding to the rth receiving antenna is equivalent to a kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna minus the kth subchannel interference reconstruction signal SICI,r,k(i).

15. The receiver according to claim 12, wherein the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna has a memory, and the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna stores the subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k))N(i−1), Ŝt,((k+1))N(i−1),..., Ŝt,((k+2))N(i−2),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1) into the memory, transmits the subchannel frequency-domain estimation signals Ŝt((k−i+2))N(1), Ŝt((k−i+3))N(2),..., Ŝt,((k−1))N(i−2), Ŝ((k))N(i−1) to the (k+1)th subchannel interference reconstruction unit corresponding to the rth receiving antenna, and transmits the subchannel frequency-domain estimation signals Ŝt((k+i−2))N(1), Ŝt((k+i−3))N(2),..., Ŝt,((k+1))N(i−2), Ŝ((k))N(i−1) to the (k−1)th subchannel interference reconstruction unit corresponding to the rth receiving antenna.

16. The receiver according to claim 12 further comprising a channel estimator, wherein the channel estimator estimates a plurality of channel impulse responses of a wireless channel and calculates a plurality of coefficients according to the channel impulse responses, wherein the kth subchannel interference reconstruction unit corresponding to the rth receiving antenna generates a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna according to (2×i−2) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+1))N(i−1), Ŝt,((k+2))N(i−2),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1), and the coefficients during the ith iteration.

17. The receiver according to claim 12, wherein the MIMO detector is a MIMO detector using a MPA.

18. The receiver according to claim 12, wherein the MIMO detector performs a plurality of sub iterations during the ith iteration, and during each of the sub iterations, the MIMO detector generates a probability message for the bit bn according to the subchannel frequency-domain signal Y′r,k(i) and then generates a probability message for the subchannel frequency-domain signal Y′p,k(i) according to the bit ba, and during the last sub iteration, the MIMO detector generates probabilities of the bits b0, b1,..., and bMT−1 according to the probability messages, wherein p is an integer between 1 and T, and a and n are integers between 0 and (MT−1).

19. The receiver according to claim 18, wherein if the ith iteration is the last iteration, the MIMO detector makes a hard decision according to the probabilities of the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1, and if the ith iteration is not the last iteration, the MIMO detector makes a soft decision according to the probabilities of the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1.

20. The receiver according to claim 12 further comprising a channel decoder coupled to the MIMO detector, wherein the channel decoder determines the subchannel frequency-domain estimation signal Ŝt,k(i) according to the bits bM(t−1), bM(t−1)+1,..., and bMt−1 of the kth subchannel of the tth transmitting antenna.

21. The receiver according to claim 12, wherein the MIMO detector determines the subchannel frequency-domain estimation signal Ŝt,k(i) according to the bits bM(t−1), bM(t−1)+1,..., and bMt−1 of the kth subchannel of the tth transmitting antenna.

22. A progressive parallel interference cancellation method, adaptable to a receiver in a MIMO communication system based on an OFDM modulation technique, wherein the receiver performs a plurality of iterations, the progressive parallel interference cancellation method comprising:

generating a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna according to (2×i−1) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+2))N(i−2), Ŝt,((k+1))N(i−1),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1), during the ith iteration, wherein i is an integer greater than 1, the subchannel frequency-domain estimation signal Ŝ1,q(x) represents the qth subchannel frequency-domain estimation signal corresponding to a tth transmitting antenna generated during the xth iteration, ((•))N represents a calculation of dividing by N to obtain a remainder, N is a number of subchannels in the MIMO communication system, k is an integer between 0 and (N−1), x is an integer between 1 and (i−1), q=k−i+x or q=k+i−x; and

obtaining a kth sub channel frequency-domain signal Y′r,k(i) corresponding to the rth receiving antenna according to a kth subchannel frequency-domain reception signal Yr,k and the kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna during the ith iteration.

23. The progressive parallel interference cancellation method according to claim 22 further comprising:

during the first iteration, presetting a kth subchannel interference reconstruction signal SICI,r,k(1) corresponding to the rth receiving antenna to 0, and obtaining a kth subchannel frequency-domain signal Y′r,k(1) corresponding to the rth receiving antenna according to the kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna.

24. The progressive parallel interference cancellation method according to claim 22, wherein a kth subchannel frequency-domain signal Y′r,k(i) corresponding to the rth receiving antenna is equivalent to a kth subchannel frequency-domain reception signal Yr,k corresponding to the rth receiving antenna minus the kth subchannel interference reconstruction signal SICI,r,k(i).

25. The progressive parallel interference cancellation method according to claim 22 further comprising:

estimating a plurality of channel impulse responses of a wireless channel, and calculating a plurality of coefficient according to the channel impulse responses,

wherein a kth subchannel interference reconstruction signal SICI,r,k(i) corresponding to the rth receiving antenna is generated according to (2×i−2) subchannel frequency-domain estimation signals Ŝt,((k−i+1))N(1), Ŝt,((k−i+2))N(2),..., Ŝt,((k−2))N(i−2), Ŝt,((k−1))N(i−1), Ŝt,((k+2))N(i−2), Ŝt,((k+1))N(i−1),..., Ŝt,((k+i−2))N(2), Ŝt,((k+i−1))N(1) and the coefficients during the ith iteration.

26. The progressive parallel interference cancellation method according to claim 22 further comprising:

determining a plurality of bits b0, b1,..., and bMT−1 corresponding to the kth subchannels of the transmitting antennas according to the kth subchannel frequency-domain signals Y′1,k(i), Y′2,k(i),..., and Y′R,k(i) of the receiving antennas during the ith iteration, wherein M is a bit modulation order, R is a number of the receiving antennas, T is a number of the transmitting antennas, and the bits bM(t−1), bM(t−1)+1,..., and bMt−1 corresponding to the kth subchannel of the tth transmitting antenna are used for determining the subchannel frequency-domain estimation signal Ŝt,k(i).

27. The progressive parallel interference cancellation method according to claim 26, wherein a plurality of sub iterations is performed during the ith iteration, and during each of the sub iteration, a probability message is generated for the bit bn according to the subchannel frequency-domain signal Y′r,k(i), and a probability message is then generated according to the bit ba for the subchannel frequency-domain signal Y′p,k(i), and during the last sub iteration, probabilities of the bits b0, b1,..., and bMT−1 are generated according to the probability messages, wherein p is an integer between 1 and T, and a and n are integers between 0 and (MT−1).

28. The progressive parallel interference cancellation method according to claim 27, wherein if the ith iteration is the last iteration, a hard decision is made according to the probabilities of the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1, and if the ith iteration is not the last sub iteration, a soft decision is made according to the bits b0, b1,..., and bMT−1 to determine the bits b0, b1,..., and bMT−1.

29. The progressive parallel interference cancellation method according to claim 26 further comprising:

determining the subchannel frequency-domain estimation signal Ŝt,k(i) according to the bits bM(t−1), bM(t−1)+1,..., and bMt−1 of the kth subchannel of the tth transmitting antenna.